BACKGROUND: Racial disparities have historically existed regarding HIV care outcomes, including linkage to care. This study aims to explore the contribution of contextual features (e.g., socioeconomic and structural environmental factors) to the temporal change of county-level racial disparities in linkage to care. METHODS: This is a statewide population-based retrospective cohort study. The patient-level variables in the South Carolina HIV registry system were used to calculate the aggregated county-level linkage to care percentage. Then, we used four indices to measure racial disparities in the county-level percentage of timely linkage to care, i.e., the Black-White ratio, index of disparity (ID), weighted ID, and Gini coefficient. Linear mixed-effect models were used to estimate the relationship between a variety of contextual features and disparity indexes. The analysis included data from 2013 to 2020, with 2013 as the start year due to the availability of key contextual features and 2020 as the end year based on the most recent HIV registry data available at the time of this study. RESULTS: Across 46 counties in South Carolina, racial disparity in linkage to care persisted between 2013 and 2020, as indicated by all four indices. When using ID, weighted ID, and Gini as outcomes, counties with lower degrees of racial residential segregation and stronger family structure were at higher risk of racial disparities in linkage to care. For weighted ID only, counties with fewer primary care providers (β = - 4.22
95% CI, - 7.23 ~ 1.23) had larger racial disparities in linkage to care. Furthermore, for Gini only, counties with higher income inequalities (β = 0.01
95% CI, 0.00 ~ 0.02) had larger racial disparities in linkage to care. CONCLUSION: Efforts to address racial disparities should continue, and innovative approaches, specifically those that focus on social and structural factors, should be developed and implemented for populations that have poor HIV outcomes in the USA.